U.S. patent number 5,133,349 [Application Number 07/573,032] was granted by the patent office on 1992-07-28 for method for adapting the stimulation frequency of a heart pacemaker to the burden of the patient.
This patent grant is currently assigned to Siemens Aktiengesellschaft. Invention is credited to Roland Heinze.
United States Patent |
5,133,349 |
Heinze |
July 28, 1992 |
Method for adapting the stimulation frequency of a heart pacemaker
to the burden of the patient
Abstract
During optimization intervals prescribed by the heart pacemaker,
the stimulation frequency is periodically changed and an
acquisition of a measured value that is dependent on the cardiac
minute output is respectively phase-synchronized with the change in
stimulation frequency. An identification is thereby made whether
the measured value that is dependent on the cardiac minute output
changes given the change of the stimulation frequency. A number of
curves are stored representing different relationships between a
regulating variable, derived from the measured value, and the
stimulation frequency. The stimulation frequency of the pacemaker
is set based on one of these curves, as selected by an optimization
controller to which the regulating variable is supplied. If the
change in the measured valued as acquired during the optimization
intervals indicates that the currently-selected curve is no longer
accurate, the optimization controller selects another curve from
among the stored curves which best represents the current
relationship between the regulating variable and the stimulation
frequency.
Inventors: |
Heinze; Roland (Munchen,
DE) |
Assignee: |
Siemens Aktiengesellschaft (Mun
chen, DE)
|
Family
ID: |
6346713 |
Appl.
No.: |
07/573,032 |
Filed: |
August 6, 1990 |
PCT
Filed: |
February 06, 1989 |
PCT No.: |
PCT/EP89/00108 |
371
Date: |
August 06, 1990 |
102(e)
Date: |
August 06, 1990 |
PCT
Pub. No.: |
WO89/06990 |
PCT
Pub. Date: |
August 10, 1989 |
Foreign Application Priority Data
Current U.S.
Class: |
607/22 |
Current CPC
Class: |
A61N
1/36557 (20130101) |
Current International
Class: |
A61N
1/365 (20060101); A61N 001/365 () |
Field of
Search: |
;128/419PG |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
Primary Examiner: Kamm; William E.
Attorney, Agent or Firm: Hill, Van Santen, Steadman &
Simpson
Claims
I claim as my invention:
1. In a method for adapting the stimulation frequency of a heart
pacemaker to the burden of a patient, including the steps of
acquiring a measured value dependent on the cardiac minute output,
changing the stimulation frequency during defined time intervals
dependent on whether said measured value dependent on the cardiac
minute volume changes, deriving a regulating variable from a
succession of said measured values, and selecting a characteristic
that best represents the relationship between the regulating
variable and stimulation frequency in accord with this measured
value change from among a plurality of stored characteristics
respectively representing different relationships between the
regulating variable and stimulation frequency, the improvement
comprising the steps of identifying a series of chronologically
spaced optimization intervals, periodically changing the
stimulation frequency during each optimization interval, acquiring
a series of said measured values dependent on the cardiac minute
output respectively at times which are phase-synchronized with the
change in stimulation frequency, and updating the selection of said
characteristic that best represents the relationship between the
regulating variable and stimulation frequency using said measured
values acquired during said optimization interval.
2. The improvement of claim 1, comprising the additional steps of
obtaining a correlation value identifying a correlation between all
measured values that are acquired during a frequency change period
and measured values of the preceding frequency change periods; and
setting the characteristics burden-dependent regulating
variable-stimulation frequency based on said correlation value.
3. The improvement of claim 1, comprising the additional steps of
calculating the measured value changes obtained during an
optimization interval, and selecting the characteristic
representing the relationship between regulating variable and
stimulation frequency based on the calculation.
4. The improvement of claim 3, comprising the additional step of
comparing the calculation of the measured value changes in an
optimization interval to an upper, positive limit value and to a
lower, negative limit value, and wherein the step of selecting the
characteristic representing the relation between regulating
variable and stimulation frequency is defined by the steps of:
a) shifting the characteristic toward lower stimulation frequencies
when the calculation of the measured value changes is less than -A1
in at least one optimization interval;
b) maintaining the characteristic unaltered when the calculation of
the measured value changes lies between -A1 and +A1; and
c) shifting the characteristic in the direction of higher
stimulation frequencies when the calculation of the measured value
changes is greater than A1 in at least one optimization
interval.
5. The improvement of claim 4, wherein the step (a) is further
defined as shifting the characteristic in the direction of lower
stimulation frequencies when the calculation of the measured value
changes lies between -A1 and A1 in two successive optimization
intervals.
6. The improvement of claim 4, wherein step (c) is further defined
as shifting the characteristic in the direction of higher
stimulation frequencies when the calculation of the measured value
changes in two successive optimization intervals lies between -A1
and A1 and a lower limit characteristic is reached.
7. The improvement of claim 6 wherein said optimization intervals
include a resting phase and a high stress phase, and comprising the
additional steps of conducting steps (a), (b) and (c) only during
said rest phase.
8. The improvement of claim 3, wherein said stimulation frequency
has an upper limit f.sub.max, wherein said optimization intervals
include a high stress phase, and comprising the additional steps of
comparing the calculation of the measured value changes to an
upper, positive limit value A1 and to a lower, negative limit value
-A1 and varying the upper limit f.sub.max according to the
following steps during a high stress phase of the patient;
a) reducing the limit value f.sub.max when the calculation of the
measured value changes .DELTA.M in at least one optimization
interval is less than -A1 and the upper limit f.sub.max has not yet
reached a lowest prescribed value;
b) maintaining the upper limit f.sub.max unaltered when the
calculation of the measured value changes .DELTA.M lies between -A1
and +A1; and
c) increasing the upper limit value f.sub.max when the calculation
of the measured value changes .DELTA.M in at least one optimization
interval is greater than A1 and the upper limit f.sub.max has not
yet reached an uppermost, prescribed value.
9. The improvement of claim 8, wherein step (a) is further defined
by reducing the upper limit f.sub.max when the calculation of the
measured value changes .DELTA.M lies between -A1 and A1 in two
successive optimization intervals and the upper limit f.sub.max has
not yet reached said lowest, prescribed value.
10. The improvement of 8, the upper limit value f.sub.max when the
calculation of the measured value changes lies between -A1 and A1
in two successive optimization intervals and the upper limit
f.sub.max has reached said lowest, prescribed value.
11. The improvement of claim 1, comprising the additional steps of
storing, for each current measured value Mn, at least two measured
values M.sub.n-1 and M.sub.n-2 acquired in preceding stimulation
frequency changes, and calculating a measured value change .DELTA.M
according to one of the following equations:
given a decrease of the stimulation frequency
12. The improvement of claim 1 comprising the additional steps of
determining whether each stimulation frequency change is positive
or negative, assigning a positive value to a measured value change
occurring as a result of a positive stimulation frequency change,
and assigning a negative value to a measured value change occurring
as a result of a negative stimulation frequency change.
13. The improvement of claim 1, comprising the additional step of,
if a number of measured value changes adequate for an optimization
procedure cannot be acquired in a current optimization interval,
storing and re-using the measured value changes from a preceding
optimization interval as measured value changes in said current
optimization interval.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention is directed to a method for adapting the stimulation
frequency of a heart pacemaker to the burden of a patient, whereby
a characteristics regulator controls the stimulation frequency
dependent on a burden-dependent regulating variable and whereby, by
changing the stimulation frequency during defined time intervals,
an identification is made to see whether a measured value dependent
on the cardiac minute output thereby changes, and whereby an
optimization controller influences a characteristic that represents
the relationship between the regulating variable and the
stimulation frequency which correspondingly influences this change
in measured value.
2. Description of the Prior Art
A method of this species is disclosed by EP-A-2 0 165 566. The
stimulation frequency of a heart pacemaker is thereby regulated
dependent on the central venous blood oxygen saturation in the
heart. The central venous blood oxygen saturation is calculated
according to the principle of reflection oximetry that, for
example, is disclosed in detail in DE-A-31 52 963.
An optimum adaptation to the individual hemodynamic situation of
the patient that changes over time is not possible with a
permanently prescribed, invariable characteristic such as the
relationship between blood oxygen saturation and stimulation
frequency. In the EP-A2-0 165 566, an optimizing control that acts
continuously in addition to the characteristics control is thereby
provided. The optimizing control continuously monitors the tendency
of the value of the blood oxygen saturation and, by automatic
elevation or lowering of the frequency, determines whether an
improvement (measured value increase) of the value of the blood
oxygen saturation ensues or not. It is not only one defined
characteristic but an entire family of characteristics that is now
allocated to every value of blood oxygen saturation.
This arrangement, however, has the disadvantage that it makes a
decision after every change in frequency whether a positive or
negative reaction of the blood oxygen saturation ensued and, thus,
there is the risk of incorrect decision as a consequence of
disturbing changes of the oxygen saturation that are dependent on
burden above all.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to design a
method of the species initially cited such that the optimum
adaptation of the regulation of the stimulation frequency to the
hemodynamic situation of the patient becomes more independent o
disturbing influences and is thus further improved.
This object is inventively achieved in that the change of the
stimulation frequency is periodically repeated during optimization
intervals prescribed by the heart pacemaker and an acquisition of
the measured value dependent on cardiac minute output respectively
ensues phase-synchronized with the change in stimulation frequency.
Disturbing influences can be largely eliminated due to the multiple
repetition of changes in the stimulation frequency.
DESCRIPTION OF THE DRAWINGS
FIG. 1 generally shows a heart pacemaker and lead in relation to a
human heart.
FIG. 2 is a graph showing the relationship between stimulation
frequency and cardiac minute output.
FIG. 3 is a schematic block diagram of the control path used in the
method of the present invention.
FIG. 4 is a schematic block diagram of a control circuit for
practicing the method of the present invention.
FIG. 5 is a graph showing a family of characteristic curves for
oxygen saturation versus stimulation frequency in accordance with
the principles of the present invention.
FIG. 6 shows a stimulation phase diagram for illustrating the
principles of the method in accordance with the principles of the
present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENT
FIG. 1 shows an arrangement of a heart pacemaker known from DE-A-31
52 963. The heart pacemaker HS contains a power supply part Ba, an
electronic switch Sch having a measuring amplifier MV, and the
two-pole electrical coupling IK. The two-pole electrical plug IS of
the stimulation catheter K is screwed fast in the coupling IK. Via
the upper hollow vein HV, the stimulation catheter K leads into the
right auricle RV in which the neutral sensing electrode E.sub.2 is
positioned and then leads into the right ventricle RHK. In the
region of the right ventricle RHK, the stimulation catheter K
contains a measuring probe MS that measures the blood oxygen
saturation according to the principle of reflection oximetry. The
cardiac muscle H is stimulated by the stimulation electrode
E.sub.1, whereby the stimulation frequency is dependent on the
measured signal of the measuring probe MS, i.e. on the blood oxygen
saturation. The blood oxygen saturation thus acts as a regulating
quantity for the control of the stimulation frequency.
In the ideal case, the heart pacemaker should set the stimulation
frequency such that the venous blood oxygen saturation is optimally
high at an optimally low frequency given a constant physical
exertion. This objective derives from the fact that the
physiological adaptation of the cardiac muscle to physical
exertions P is optimum when the cardiac minute output HMV is
proportional to the physical exertion P: HMV=P. In order to
optimize the burden of the cardiac muscle, a defined cardiac minute
output with optimally high stroke volume SV.sub.max should be
achieved given optimally low pulse frequency f.sub.bmin, i.e.
It is also true that given a constant physical exertion
(P=constant), i.e. given a constant oxygen drain of the blood in
the peripheral circulation, the change of the venous blood oxygen
.DELTA.SO.sub.2 is proportional to the change of the cardiac minute
output .DELTA.HMV. On the basis of the aforementioned
relationships, the heart pacemaker--on the basis of an optimum
value regulation--should select the characteristic curve from a
family of characteristic curves which relate the cardiac minute
volume (or a variable derived therefrom) to the--stimulation
frequency that is best adapted to the efficiency of the cardiac
muscle.
FIG. 2 shows the relationship between the heart rate f.sub.p and
the cardiac minute output HMV for a healthy myocardium (a) and for
two cases of a diseased myocardium (curves b, c). One can see from
the curves that the cardiac minute output HMV in the low load
region increases approximately proportionally to the heart rate fp
and, dependent on the efficiency of the myocardium, asymptotically
reaches a limit value HMV.sub.max. Characteristic of the diseased
myocardium is the drop of the cardiac minute output HMV beginning
with a maximum heart rate f.sub.pmax, for which reason a heart rate
above f.sub.pmax as stimulation frequency is to be avoided in all
cases.
FIG. 3 shows a control circuit for a physiological regulation of
the stimulation frequency. The heart H is thereby to be considered
as a control element that acts on the blood circulation BL via the
cardiac minute output HMV. The blood circulation BL is also
influenced by the physical exertion P that is to be considered as a
disturbing quantity.
A first sensor S1 is provided for the acquisition of a
burden-dependent physiological quantity, for example EKG, blood
pressure, respiratory volume, or of a measured activity value of
the physical exertion. Dependent on the required physiological
quantity, the sensor S1 is interactively connected for this purpose
to the heart H, to the blood circulation BL or to a quantity (P)
that represents the physical exertion P, for example, the activity.
Via a substraction element whose second input is supplied with a
first reference variable input FG1, a regulating variable is formed
from the output signal of the sensor S1, this regulating variable
prescribing the stimulation frequency f for the heart H via the
controller 14.
As a regulating signal, an optimization controller 11 receives the
cardiac minute output HMV acquired via a second sensor S2, this
being subtracted from a second reference variable input FG2. The
sensor S2, for example, can thereby work according to the principle
of impedance measurement. The optimization controller 11 regulates
the relationship between the regulating variable and stimulation
frequency prescribed by the controller 14 to an optimum value.
As shown in FIG. 4 another possibility for optimum regulation uses
only one measuring probe MS for the blood oxygen saturation
SO.sub.2. The measured value of the blood oxygen saturation
SO.sub.2 is thereby supplied both to the controller 14 as well as
to the optimization controller 11, being supplied thereto as the
regulating variable. The sensors S1 and S2 are eliminated. In the
embodiment of FIG. 3, FG1 is a reference variable for the first
physiological quantity provided by sensor S1, and FG2 is a
reference variable for the second physiological quantity provided
by sensor S2. In the embodiment of FIG. 4, the selection of a curve
corresponds to the reference variable FG2, and the curve itself
corresponds to FG1.
The measuring probe MS is thereby connected to a measured value
memory 4 via a switch 1, a measuring amplifier 2 and an averaging
unit 3. The measured value memory 4 contains memories 4a-4c for the
respectively current measured value M.sub.n and two respectively
preceding measured values M.sub.n=1 and M.sub.n-2.
Two subtraction elements 5 and 6 follow the measured value memory
4, whereby the subtraction element 5 subtracts the measured value
M.sub.n-2 from the measured value M.sub.n and the subtraction
element 6 subtracts the measured value M.sub.n-2 from the measured
value M.sub.n-1. The output of the subtraction element 5 is
connected via a proportional element 5a having the proportional
factor 1/2 to the minus input of a further subtraction element 7
and the subtraction element 6 is connected to the plus input of the
subtraction element 7. Via a switch 8, the subtraction element 7 is
optionally connectible directly to the input of a summing element
10 for the measured values obtained in an optimization phase (yet
to be set forth), via an inverter 9 to the same summing element 10
or to a free contact. The switch 8 is driven by an optimization
controller 11 acting as an optimum regulator. The output of the
summing element 10 is connected via a comparator 12 to an input of
the optimization controller 11. Further, the optimization
controller 11 is controlled by the measured value taken at the
output of the averaging unit 3. The optimization controller 11 acts
on a controller 14 having a curve generator. Dependent on the value
curve prescribed by the optimization controller 11, the controller
14 regulates the frequency f of a pulse generator 15 that is
connected to the stimulation electrode E1.
The functioning of the control circuit shall be set forth in
greater detail below with reference to FIGS. 5 and 6.
FIG. 5 shows a family of curves for the relationship between the
normed blood oxygen saturation SO.sub.2norm and the stimulation
frequency f. The blood oxygen saturation SO.sub.2 is acquired with
the measuring probe MS, is amplified with the measured value
amplifier 2, is averaged with the averaging unit 3 and is normed in
the optimization controller 11. The curves K0-K4 are stored in the
curve generator of the controller 14, whereby switching to
different curves can be undertaken based on a signal from
optimization controller 11.
Further, five maximum values f.sub.max0 -f.sub.max4 for the
stimulation frequency are defined. The maximum values f.sub.max0
-f.sub.max4 are defined by the optimization controller 11. Each
curve K0-K4 can be cut by one of the maximum values f.sub.max0
-f.sub.max4. A curve K that is best adapted to the efficiency of
the cardiac muscle is to be selected with the optimum regulation
from the family of characteristics shown in FIG. 5.
To this end, two characteristic, physical conditions or phases are
first defined wherein an individual adaptation of the stimulation
frequency is particularly important, namely, in the resting phase
and in the high-stress phase. Both phases are shown in FIG. 5.
In the "resting phase", the physical activity acquired by a region
of high, normed oxygen saturation SO.sub.2 is reduced to a minimum.
A low oxygen drain is present and the central venous value of the
oxygen saturation reaches its maximum.
In the "high-stress phase", the physical performance acquired by a
range of low, normed oxygen saturation SO.sub.2norm reaches an
individually typical, maximum quantity, i.e. a maximum oxygen drain
is present given a minimum central venous value of the oxygen
saturation SO.sub.2.
It should be noted that the terms "in phase" and
"phase-synchronized" as used herein have their normally understood
wave mechanics meaning of two signals having the same phase angle
or phase angles differing by a fixed amount. These terms do not
refer to the above "resting phase" or "high-stress phase".
Ideal conditions for an optimizing frequency adaptation would be
established if a central venous oxygen saturation SO.sub.2 in the
range between 70 and 80%, i.e. a value as in the case of a person
with a healthy heart, were to be achieved and if a cardiac minute
output that is maximum for the patient could be set in the
high-stress phase.
The prerequisite for such an optimization, however, would be an
absolute measurement value of the oxygen and of the stroke volume.
These measured quantities, however, cannot be directly acquired
with existing methods. A measuring method wherein the measured
quantities are indirectly acquired is therefore used for the
optimum regulation.
The optimization method employed is fundamentally based, due to the
periodic change of the stimulation frequency and the in-phase
measured value calculation, on the acquisition of data identifying
as to whether and how the frequency change influences the
hemodynamic situation, i.e. the cardiac minute output of the
patient. Since the cardiac minute output cannot be directly
acquired, the oxygen saturation SO.sub.2 dependent on the cardiac
minute output is employed as an equivalent quantity.
Insofar as the optimization circuit is activated, optimization
intervals are carried out at regular intervals of, for example, six
hours initiated by a timer 11a in the optimization controller 11.
An optimization is thereby carried out only during the resting
phase and during the high-stress phase. The sequence of an
optimization shall be set forth in greater detail below with
reference to the diagram of FIG. 6.
FIG. 6 shows an assumed burden curve P of the patient for an
optimization interval I (resting phase) and for an optimization
interval II (high-stress phase) that lies a number of hours later.
The stimulation frequency f that derives on the basis of a
preselected curve K is shown with dotted lines over the burden P.
Further, the curve of the oxygen saturation SO, is shown, whereby
the dotted line again represents the curve deriving without optimum
regulation. The optimization in the resting phase I starts with a
lowering of the stimulation frequency f, namely, by, for example,
eight beats per minute in the resting phase. This lowering of the
frequency lasts, for example, 48 heartbeats and is subsequently
cancelled for 48 heartbeats and is then repeated. In order to
suppress brief-duration disturbing influences--for instance, due to
burden fluctuations--16 frequency changes, for example, are
executed per optimization interval and disturbing influences are
then eliminated by formation of an average signal, or a combination
of signals over them, according to known methods.
As the pulse diagram of FIG. 6 shows, a measured value M.sub.n of
the oxygen saturation SO.sub.2 is stored synchronously with every
frequency change. The measured values M.sub.n are measured with a
time delay (hysteresis) of four beats in order to take the delay of
the blood stream in the venous circulation into consideration.
Two preceding measured values M.sub.n-1 and M.sub.n-2 are stored in
the measured value memory 4 for every measured value M.sub.n.
The calculation of the measured values .DELTA.M ensues after every
change period, i.e., for example, after 96 heartbeats, whereby the
measured value difference .DELTA.Mn that is determined is defined
according to the following equation:
If the difference between two successive measured values were
simply employed, then the influence of burden changes that yield
the dotted curve of the oxygen saturation SO.sub.2 would thus also
have been acquired. However, only the influence of the frequency
change is to be identified. When it is assumed that the curve of
the oxygen saturation SO.sub.2 is linear without the periodic
change in stimulation frequency, then a change of the oxygen
saturation of 1/2 (M.sub.n -M.sub.n-2) that is independent of the
periodic frequency change is obtained at the measured value
M.sub.n-1. When, given an elevation of the stimulation frequency,
this factor 1/2 (M.sub.n -M.sub.n-2) is subtracted from the overall
measured value change (M.sub.n -M.sub.n-2) or, respectively, this
value is added given a lowering of the stimulation frequency f,
then the change .DELTA.M of the measured value for the oxygen
saturation caused by the periodic change of the stimulation
frequency is obtained with a good approximation.
As already mentioned, 16 difference measurements, for example, are
preferably executed in succession, whereby only one part of the
frequency change is shown in FIG. 6 for the sake of clarity. A
combination of all 16 measured value changes .DELTA.M is then
utilized for the optimization regulation.
For example, a correlation function known from Profos, Handbuch der
industriellen Messtechnik, 1978, Pages 185 through 189 could
thereby be utilized. In the described exemplary embodiment,
however, the sum .SIGMA..DELTA.M of the measured value changes is
evaluated in that the measured values .DELTA.M pending at the
output of the summing element 7 are supplied to the summing element
10.
In the sum formation of the measured value changes .DELTA.M, of
course, the measured value changes .DELTA.M obtained during
frequency lowerings must be evaluated with the inverse operational
sign compared to the measured value change .DELTA.M obtained given
frequency elevations, this ensuing with the inverter 9.
The calculation of the measured value changes in a high-stress
phase is implemented in analogous fashion during an optimization
interval II. In the high-stress phase, the change of the
stimulation frequency with, for example, 16 per minute can be
selected higher than in the resting phase. A measured value change
.DELTA.M employed for the optimization regulation is also obtained
in the high-stress phase by sum formation after 16 different
measurements.
It is possible that the provided number of measured value changes
.DELTA.M, 16 in the example, cannot be acquired in a respective
optimization cycle I II because, for example, the burden P of the
patient has changed in the meantime. In this case, the measured
value changes .DELTA.M that are already acquired are stored and are
used for the next optimization cycle.
The measured value .SIGMA..DELTA.M pending at the output of the
summing element 10 after the implementation of the measured value
acquisition addressed above are now compared to two thresholds
-A.sub.1 and +A.sub.1 in the comparator 12. Dependent on the result
of the comparison, the optimum curve is selected from the family of
curve of FIG. 5 in accordance with the table on page 13. In the
resting phase, it is thereby only the curve K that is defined,
whereas the maximum stimulation frequency f.sub.max is determined
in the high-stress phase. When it is assumed that the
characteristic 2 shown emphasized in FIG. 5 was first set in the
curves generator of the controller 14 and the maximum stimulation
frequency f.sub.max2 was set, then, for example, it is thus the
curve K1 that is selected when the sum of the measured value
changes .SIGMA..DELTA.M in at least one optimization interval is
<A.sub.1, i.e. when an improvement of the oxygen saturation was
thus achieved by reducing the stimulation frequency f. By contrast,
the curve K3 is selected when the sum of the measured value changes
in an optimization interval I, i.e. in the resting phase, is
>-A.sub.1. This means that an improvement of the oxygen
saturation SO.sub.2 was achieved by elevating the stimulation
frequency, so that the curves K3 seems better-suited for the
patient.
When the sum of the measured value changes lies between -A.sub.1
and +A.sub.1 in an optimization interval I in the resting phase, it
can be concluded that a change of the stimulation frequency has no
significant influence on the blood oxygen saturation SO.sub.2. The
curve K2 is thus initially retained. When, however, the sum of the
measured value changes .SIGMA..DELTA.M in two successive
optimization intervals I lies between -A.sub.1 and +A.sub.1, the
characteristic K1 is activated. This derives from the efforts to
always give lower stimulation frequencies precedence in doubtful
cases since the cardiac muscle is thereby treated more gently. If a
deterioration of the oxygen saturation were to result from the
switch to the curve K1, then the sum of the measured value changes
.SIGMA..DELTA.N will again be >-A.sub.1 in the next optimization
interval I and the curve K2 will thus be again selected.
When, however, the characteristic K0, i.e. the lower limit curve,
is activated and the sum of the measured value changes
.SIGMA..DELTA.M lies between -A.sub.1 and +A.sub.1 in two
successive optimization intervals I, a switch to curve K1 is
undertaken. This prevents the optimization regulation from getting
stuck at the lower limit curve K0.
Analogous measures according to the illustrated table are also
implemented in the case of an optimization interval II for the
high-stress phase. The only difference that exists is that it is
not the curves K per se that are thereby influenced, but the limit
value f.sub.max for the stimulation frequency f.
Various values for the resting phase and for the high-stress phase
can be prescribed for the constants A1 and A2 dependent on the
sensitivity of the measuring probe MS and on the hemodynamic
reserve of the patient.
__________________________________________________________________________
Set Characteristic or, respectively, Change Given Combination
Condition limit frequency Resting Phase High-Stress Phase
__________________________________________________________________________
1 1 time < A.sub.1 1-4 Lower to the next curve Lower the limit
frequency by one (K0-K3) step (f.sub.max 0 -f.sub.max 3) 2 1 time
< A.sub.1 0-4 Curve Remains Limit frequency remains and >
-A.sub.1 3 2 time < A.sub.1 1-4 Lower to next curve Lower limit
frequency by one and > -A.sub.1 (K0-K3) step (f.sub.max 0
-f.sub.max 3) 4 2 time < A.sub.1 0 Increased to next curve
Increase limit frequency by one and > -A.sub.1 (K1) step
(f.sub.max 1) 5 1 time > -A.sub.1 0-3 Increase to curve Increase
limit frequency by one (K1-K4) step (f.sub.max a -f.sub.max
__________________________________________________________________________
4)
* * * * *